Visual-Based Deficits

Reversal Errors

Over the years, much attention has been focused on the reversal errors made by children with RD. These errors, which involve, for example, the reading/writing of b for d or was for saw, have traditionally been linked closely with dyslexia. Even today, most people still think of dyslexia as a problem reading letters or words backwards. Despite this view, there is surprisingly little research that has systematically investigated reversal errors. The few studies that have examined reversal errors have found that these errors do not actually occur that often in children with RD (Fischer, Liberman, & Shankweiler, 1978; Liberman, Shankweiler, Orlando, Harris, & Berti, 1971). Furthermore, when considered in terms of percentage of overall errors, reversal errors may be no more prevalent in young poor readers than they are in young good readers (Holmes & Peper, 1977). In other words, all beginning readers occasionally make reversal errors, just as all children learning to talk make errors involving grammatical morphemes (e.g., past tense -ed, third person -s). Just as children with language delays continue to have difficulty with grammatical morphemes beyond the developmental period, children with RD often continue to make reversal errors in later grades.

When reversal errors do occur, they generally are not the result of perceptual problems. Children who write saw as was or girl as gril typically do not have trouble perceiving letter sequences. Vellutino and his colleagues (Vellutino, Pruzek, Steger, & Meshoulam, 1973; Vellutino, Steger, DeSetto, & Phillips, 1975) found that children with RD could accurately copy what they sometimes failed to read correctly. Rather than having problems perceiving letter sequences, poor readers more likely have difficulties remembering the order of letters in words. Because of the spatial orientation of words, a primary way a word can be misspelled! misread is to fail to remember the correct order of its letters.

Visual Memory

Apparent problems in the memory for the letters in words led some early investigators to propose that poor readers had generalized deficits in visual memory (Fildes, 1922). Vellutino (1979), however, maintained that most of the early work showing deficits in visual memory was confounded by the use of stimuli that could be verbally labeled. Consequently, children with RD might have performed poorly because of verbal memory deficits rather than visual deficits. In support of this possibility, Vellutino and his colleagues (Vellutino et al., 1975) showed that poor readers scored comparably to good readers on a visual memory task involving stimuli that could not be easily labeled (but see Willows, Kruk, & Corcos, 1993).

Rather than focusing on a generalized problem in visual memory, some researchers have investigated the possibility that poor readers have specific problems in orthographic processing. Orthographic knowledge involves the knowledge of letter sequences or spelling patterns. This knowledge allows the reader to directly access semantic memory without going through the intermediate step of phonological decoding. Orthographic processing has often been tested by tasks that ask subjects to choose which of two letter sequences (goat, gote) is a real word. Because the foil in each word pair (gote) can be pronounced like a real word, the subject must rely on orthographic knowledge to answer correctly. Research using this task has shown that orthographic processing ability is related to reading achievement in that children with good orthographic knowledge read better than those with limited orthographic knowledge (Conners & Olson, 1990; Stanovich & West, 1989). Researchers have been quick to point out, however, that orthographic processing skills may be heavily influenced by phonological processing abilities (Share & Stanovich, 1995). Children who have mastered the use of sound-letter correspondence rules should develop richer orthographic knowledge by virtue of many successful trials reading words. Nevertheless, some studies show that orthographic processing may make an independent contribution to reading ability (Barker, Torgesen, & Wagner, 1992; Conners & Olson, 1990; Stanovich & West, 1989). Such findings suggest the possibility that some children with RD may have specific deficits in remembering the letters in words.

Erratic Eye Movements

When reading, we get the impression that our eyes are moving smoothly and continuously across the printed page. Actually, eye movements for reading (and many other visual activities) involve a series of rapid jerks, called saccades, that move from left to right, and occasionally from right to left (i.e., regressions). Each of these saccades is followed by a short fixation period averaging 200 to 250 milliseconds. It is during these fixations that information is obtained for the purpose of recognizing words.

Could problems in eye movements be a cause of reading disabilities? Poor readers have been noted to have more fixations per line, longer fixations, shorter saccades, and more regressions than good readers (Rayner, 1978). Rayner (1985) and others point out, however, that these differences in eye movements may actually be a reflection of cognitive processing difficulties during reading rather than problems in oculomotor control. For example, because poor readers take longer to recognize words and often need to go back to refresh their memory, they may show longer fixations and more regressions. In opposition to such a conclusion, Pavlidis (1981, 1985) has reported that dyslexics demonstrated abnormal eye movements in non-reading tasks (also see Eden, Stein, Wood, & Wood, 1994, 1995). Olson, Conners, and Rack (1991), however, have argued that even such findings could be a consequence of a reading problem and not a cause. They demonstrated that when poor readers were matched for reading skill with younger normal readers, no differences were observed in eye movements during non-reading tasks (but see Eden et al., 1994).

The belief that erratic eye movements are a cause of reading disabilities has often led to the popularity of visually oriented treatment approaches that involve "eye movement training" devices (Metzer & Werner, 1984). The assumption is that if poor readers could learn to move their eyes in a smoother, less erratic fashion, reading would improve. But as we pointed out earlier, the basic premise that skilled reading involves smooth eye movements is false. Not surprisingly, these training programs have not proven to be effective. Today, most professionals agree that oculomotor exercises, and behavioral optometry in general, have little to offer in the treatment of reading disabilities (Clark & Uhry, 1995; Keogh & Pelland, 1985; Silver, 1995).

Scotopic Sensitivity Syndrome

In 1983, Irlen introduced a visual-perceptual condition called scotopic sensitivity syndrome (SSS) (Irlen, 1983). This condition was argued to result from an oversensitivity to particular frequencies of light. Individuals with SSS were noted to experience a variety of problems during reading, including perceptual distortions, reduced visual field, poor focus, eye strain, and/or headaches. Irlen reported that colored eyeglass lenses or tinted plastic overlays could eliminate troublesome wavelengths of light and reduce the symptoms of SSS. The use of colored lenses/overlays soon became part of a commercial enterprise. Colored lenses/overlays can now be purchased at clinics, and even through advertisements in Reading Today, a publication of the International Reading Association. Because it is often claimed in promotional materials that many dyslexics suffer from SSS, colored filters have become an alternative, but controversial, treatment for reading disabilities (Silver, 1995).

Despite heavy press coverage, supportive testimonials, and some research, little is still known about SSS and its role in reading disabilities (Stanley, 1994). As Stanley (1994) points out, the condition is probably misnamed since most reading involves the photopic, rather than the scotopic visual system. Furthermore, it is unclear what mechanisms may be responsible for the symptoms associated with SSS and how colored lenses may affect these mechanisms. Deficits in visual timing (discussed in the next section) have been linked with SSS (Breitmeyer, 1989; Weiss, 1990), but the relationship between these deficits, SSS, and improvements with the use of colored filters is far from clear (Stanley, 1994). Of more significance is the fact that there is stil1little empirical evidence to show a causal link between SSS and reading disabilities. Despite what is claimed in promotional materials and publications (Irlen & Lass, 1989), it is unclear if children with RD have a higher incidence of SSS than nondisabled readers. It is also unresolved whether SSS, if present, is a cause of reading disabilities or an associated problem.

Notwithstanding the above concerns, recent studies have begun to examine the effectiveness of colored filters. Some investigations have found significant improvements in vision and/or reading with the use of colored lenses or overlays (Fletcher & Martinez, 1994; Robinson & Conway, 1990; but see Blaskey, Scheiman, Parisi, Ciner, Gallaway, & Selznick, 1990, and Cotton & Evans, 1989). However, much of this improvement could be due to a placebo effect or an arousal effect. Weming colored glasses or using tinted overlays could motivate some poor readers to improve or could affect their mood, and thus, their performance (Cotton & Evans, 1989; Stanley, 1991, 1994). These and other problems make it difficult to recommend the use of colored lens or overlays as a viable treatment alternative for reading disabilities (Parker, 1990; Stanley, 1991).

Transient Processing Deficits

Scotopic sensitivity syndrome and problems in eye movements have both been suggested to be the result of more primary deficits in visual processing. Researchers have identified two basic visual processing systems, the transient and sustained systems (Campbell, 1974; Graham, 1980). Each system appears to specialize in the processing of particular visual information. The transient system seems to be especially sensitive to global visual features and is thought to play an important role in guiding eye movement. The sustained system, on the other hand, responds to fine detail and is used in visual feature identification (e.g., letter/word recognition). Both of these systems must operate efficiently to meet the visual perceptual demands of reading.

Lovegrove and his colleagues (Lovegrove, 1992; Lovegrove, Martin, & Slaghuis, 1986) have observed that individuals with RD have significant difficulties on a number of nonverbal visual tasks believed to involve the transient system. They proposed that individuals with RD may have a sluggish transient processing system. The slowed processing of the transient visual system could disrupt parallel operation with the sustained system, which in turn might lead to visual distortions and other visual problems during reading.

Others have also found individuals with RD to have deficits on visual tasks related to transient processing (Cestnick & Coltheart, 1999; Eden et al., 1995; Livingstone, Rosen, Drislane, & Galaburda, 1991; Solman & May, 1990). In addition, these behavioral findings are consistent with reports of recent anatomical and physiological deficits in dyslexia (Eden et al., 1996; Livingstone et al., 1991). Livingstone and colleagues (1991), for example, found in postmortem examinations that dyslexics may have less organized and smaller neurons in the brain regions associated with transient visual processing than do normal individuals. Also, as noted above, Eden and colleagues (1996) reported that dyslexics show less task-related activation in these brain regions.

Although there is some converging support of a transient visual processing deficit in poor readers, more than a few studies have failed to find evidence of these deficits (e.g., Chiappe, Stringer, Siegel, & Stanovich, 2002; Hayduck, Bruck, & Cavanagh, 1992; Hogben, Rodino, Clark, & Pratt, 1995). Some of the conflicting findings across group studies could be the result of these deficits being present only in a subset of poor readers. Thus, the subject composition of a given study could influence its outcome. Consistent with this explanation, Ramus (2003) calculated that only 29 percent of poor readers across a number of recent studies (those that presented individual subject data) had visual processing deficits. Conflicting results could also be due to methodological differences in the way visual processing has been measured across studies. Some have further argued that visual processing deficits might in part be explained by problems in attention or motivation (Stuart, McAnally, & Castles, 2001).

Regardless of the issues concerning conflicting results, the question still remains whether visual deficits, if present, are a sufficient cause of reading disabilities. Some have questioned, for example, how transient deficits themselves could lead to the range of problems seen in children with RD (Skottun, 2000). Also, at least some evidence suggests that transient processing deficits often occur in concert with phonological processing deficits (e.g., Eden et al., 1995). A visually based explanation of reading disabilities would be better supported if a group of children with RD could be identified who have a documented history of visual deficits but no impairments in phonological processing or other known causal factors (Share & Stanovich, 1995).